Development of a Direct Injection Retrofit Kit for

Reducing Emissions from 2-Stroke Cycle Engines in Asia

Dr. Bryan Willson Prof essor and Research Director
Nathan Lorenz, Tim Bauer Research Assistants
Engines and Energy Conv ersion Laboratory
Department of Mechanical Engineering Colorado State Univ ersity
Fort Collins, CO 80523-1374 USA
E-mail: Bryan.Willson@colostate.edu

ABSTRACT

There are over 50 million 2-stroke cycle engines in Asia, powering motorbikes, mopeds,
“three-wheelers”, “tuk-tu ks”, and “tricycles”. There are over 100,000 2-stroke “tricycles” in Metro
Manila used in taxi service, and over 1.7 million in the Philippines. The tricycles’ 2-stro ke engines
are characterized by high levels of hydrocarbon (HC), carbon monoxide (CO), and particulate
matter (PM) emissions. A project has been initiated in Metro Manila to demonstrate the potential
of “direct in-cylinder fuel injection” (DI) to reduce these emissions. Based on other retrofit
applications of the technology, expectations are for a 70%-90% reduction of HC, 50%-70%
reduction in CO, 80% reduction in PM, 50% reduction in oil consumption, and 35% reduction in
fuel consumption. Phase I of the project involved demonstration of a single tricycle with “retrofit
DI”. The initial demonstration was performed in Manila in November of 2003. A complete
emissions characterization of the vehicle will begin in January 2004. Phase II (2004) is a 10-15
vehicle field test in Metro Manila to document in-use emissions reductions and fuel savings.
Phase III (2005) involves the establishment of the infrastructure for widespread dissemination and
initiation of a large retrofit program. The final installed cost of the unit has been targeted between
$10,000 - $12500 Philippine Pesos ($200-$250USD). An important part of the project involves
capacity building; Colorado State University is the technical lead, but is working with NGOs,
universities, development agencies, and driver organizations in Metro Manila to develop the
necessary skills, technical capabilities and organizational infrastructure. If successful, expansion
of the project throughout the Philippines and to other Asian countries will likely occur.

Keyw ords: two stroke, direct injection, tricycle, emissions, pollution, poverty alleviation
1.0 INTRODUCTION / PROJ ECT CONCEPT

Carburetion

Air pollution is on the increase in many Asian cities due in part to the widespread use of
carbureted two-stroke cycle engines. These engines are typically used as the power source for
“two-wheelers” (i.e. motorcycles, mopeds, etc.) and “three-wheelers” such as tricycles and tuk-
tuks. The main reasons for this are the rugged construction, low cost and high power-to-weight
ratio of the two stroke engine. Unfortunately, two-stroke cycle engines are also characterized by
high levels of unburned hydrocarbons, carbon monoxide, and particulate emissions. The high
hydrocarbon emissions result from the scavenging process used by two stroke engines.
Scavenging refers to the process by which the burned exhaust gasse s are flushed from the
engine. In a conventional “carbureted” two-stroke engine the fuel is entrained in the intake air
stream before the combustion air enters the crankcase. The charge is compressed in the
crankcase by the underside of the piston, and enters the cylinder when the piston uncovers the
transfer ports. Combustion products from the previous cycle are forced or “scavenged” from the
cylinder with this new air/fuel charge. Unfortunately, the exhaust ports are also open at this time,
allowing 30%-40% of the fuel to be lost directly into the exhaust stream.1 At idle conditions the
losses can be as high as 70%. The high carbon monoxide emissions result from the unstable
combustion inherent to carbureted two stroke engines. This instability requires that the engine be
operated with rich air/fuel ratios to maintain acceptable combustion stability. This leads to
incomplete combustion and high carbon monoxide levels. Finally, the high particulate emissions
result from the unstable combustion, excessive lubrication (typical in small two stroke engines),
and a lubrication system which allows lubricating oil to be dissolved in the fuel. In a typical 2-
stro ke, the oil mixes with the fuel at the carburetor. As the air/fuel/oil mixture transfers into the
crankcase, the fuel dissolves the oil. This action reduces the amount of oil deposited on the
cylinder wall (or other critical components) as it is essentially ‘washed’ out of the engine by the
fuel.

Direct In-Cylinder Fuel Injection

Direct in-cylinder fuel injection (direct injection, DI) is a technology that has shown the ability
to greatly reduce emissions from two-stroke engines. In a DI system the carburetor is eliminated,
and the fuel is introduced into the combustion chamber via an injector mounted in the top of the
chamber’s cylinder head. This allows exhaust products to be scavenged from the cylinder using
air only. Fuel is injected into the cylinder later in the cycle, greatly reducing the amount of
unburned fuel that is allowed to escape during scavenging. The DI process allows for a locally
rich region around the spark plug, eliminating the need to enrichen the entire cylinder to achieve
stable combustion. Elimination of rich air/fuel ratios significantly reduces carbon monoxide
emissions.

Retrofit Applications

The goal of this project is to implement a widespread dissemination program which would
enable widespread conversion of carbureted two stroke engines to operation on direct injection.
The direct injection technology has been well proven in original engine manufacture (OEM)
applications, but has never been commercialized in a retrofit application. In order to successfully
implement a widespread retrofit program, a three-phase plan has been established. Phase I of
the project involves conversion of a single vehicle. This Phase I milestone was achieved on
schedule, with the successful demonstration of an “alpha” prototype DI tricycle at the Clean Air
Now conference in Manila in November 2003. The vehicle used was a Kawasaki HDIII 125 cc
motorcycle; it was selected because it is the most popular motorcycle used in tricycle service in
 Conventional Carbureted 2-Stroke Engine

Exhaust
Fuel & oil 35%-70% of
fuel los t
Intake during
scavenging

Carburettor

Direct-Injected 2-Stroke Engine

Stratified fuel
injection after
piston covers
exhaust port Exhaust Exhaust
Only air lost
during
Intake scavenging

AIR &
METERED OIL

the Philippines. Phase II involves retrofitting 10-15 motorcycles in 2004 with a near production
level “beta version” retrofit kit and testing them extensively, primarily in Metro Manila. Phase III is
the launch of the widespread retrofit program and is scheduled for 2005.

2.0 TECHNOLOGY DESCRIPTION – DIRECT INJ ECTION

In a carbureted engine the fuel is largely vaporized before entering the engine. In a DI
engine, the fuel is injected as the piston is returning upward on its compression stroke, limiting
the time for fuel to vaporize and mix into a combustible mixture. This time constraint can be
overcome by use of an injector that atomizes the fuel into very fine droplets as it is injected.
Currently there are several competing DI technologies, each characterized by a different method
of atomizing the fuel. The project team selected the Orbital air assisted fuel injection system as
most appropriate for use in retrofit applications.
Orbital Engine Corporation (OEC) has developed the air-assisted, spray guided injection
system known as the Orbital Combustion Process (OCP). 2 The OCP system utilizes a gasoline
injector to meter the fuel. Compressed air is then used to atomize the fuel into tiny droplets. This
requires the use of a second injector - the air injector - supplied by a small engine-driven air pump.
The OCP system incorporates these injectors into a single injector housing that can be seen on
the retrofitted engine in Figure 2. This system results in improved charge control, increased
combustion stability, decreased short-circuiting of fuel, and reduced knock sensitivity. The engine
control unit (ECU), a low-cost computer, controls the timing and duration of fuel and air injection.
Lubricating oil must still be metered into the engine. However, since the oil is no longer
dissolved in fuel, its behavior in the combustion process is fundamentally different, resulting in
dramatically reduced production of “white smoke”. This is demonstrated by that fact that a fuel-
to-oil ratio of 30:1 (typical of carbureted two-strokes) results in virtually smoke-free operation on a
DI two-stroke, but produces copious quantities of smoke on a carbureted two-stroke. In addition,
since the oil is not washed from internal components by the fuel, the overall oil requirement can
be reduced in a DI two-stroke. The OCP system uses a solenoid-driven oil pump controlled by
the ECU. This provides the flexibility to “map” the lubrication rate based on speed and load, thus
tailoring the oil admission to the engine’s lubrication requirements.
As the above description confirms, additional hardware is required to retrofit DI on an
existing engine. Figure 2 shows key elements of CSU’s successful retrofit kit installed on a 125
cc Kawasaki HD-III engine. A few items are not visible in the picture: the ECU, throttle body,
throttle position sensor and ignition coil.

Fuel Injector

Fuel Rail

Air Injector

Cylinder Head

Oil Pump

Fuel Pump

Air Pump

Magnetic Pick Up

Air Pump Cam

Figure 2 – Retro fit kit installed on a 125cc Kawasaki HD-III motorcycle

3.0 PROJ ECT PHASES

The structure developed for implementation of this project involves three phases. The
first phase involves building one prototype conversion kit. The second phase is a beta test of 10
to 15 units, and the final phase is widespread implementation. Phase I of the project began in
June of 2003 with the shipment of a Kawasaki HD-III to Colorado State University. There, a team
of engineers designed and manufactured the components necessary to convert the motorcycle to
direct injection. Initial calibrations are complete, and the motorcycle was demonstrated in Manila
in November, 2003. This unit is an “alpha unit”, which means that the hardware is in an early
form - not in the final embodiment envisioned for the Phase III implementation. For example, the
new head for the Phase I alpha unit was machined from a solid billet of aluminum, a prohibitively
expensive approach for widespread production.
Phase II of the project is scheduled to begin in January 2004. The equipment used in the
Phase II field test will be “beta hardware,” signifying that it is a step closer to the final production
configuration. The heads for the beta engine will be cast (in the Philippines), instead of being
machined from billet stock. Significant effort will be directed to the reliability and cost reduction of
the kit. The beta units will operate under ‘real world’ conditions to validate reliability, fuel economy,
in-use emissions, and real world performance. During this field test, the motorcycles will be
closely monitored, with particular emphasis on fuel consumption, emissions, maintenance
requirements, and driver acceptance. This data will help to finalize the design constraints used
for production design and calibration.
Phase III, currently slated for early 2005 involves the widespread implementation of direct
injection retrofit kits and continued capacity building. Results from the engineering tests in Phase
II will be implemented to increase reliability and minimize production costs. Installation centers
will expand to cover a larger geographical area in the Philippines.

4.0 EXPECTED OUTCOMES

Expected outcomes of the overall project are summarized in this section. Table 1 shows
emissions results that were achieved during investigation of a retrofit conversion project in India
by Orbital. Unburned hydrocarbons (HC’s) and carbon monoxide (CO) were significantly reduced
in the transition from carburetion to direct injection. Colorado State University has significant
experience in applying the OCP system to snowmobiles. The snowmobile application has
demonstrated 90% reduction in HC’s, 70% reduction in CO, and 35%+ reduction in fuel
consumption. The project team expects comparable results for the tricycle retrofit.

Vehicle Tested HC+NOx (g/km ) CO(g/km ) Fuel consum ption (km/l)

Carb. Baseline
150 cc scooter 3.8 1.2 50
Orbital DI
150 cc scooter 0.9 0.7 70.5
2005 India em issions
legislation proposal 1.5 0.7 NA
Table 1 - Indian motorcycle exhaust emissions results with air-assist direct
injection 3
Another expected outcome of the tricycle conversion project is for the owner/operator to
see a significant savings in fuel and 2T oil consumption. Based on work completed by both
Orbital and CSU3,4 fuel savings should be on the order of 35%, and 2T consumption should be
reduced by half. This has significant energy implications. If 1 million of the 1.7 million registered
two-stro ke tricycles in the Philippines were converted, the potential fuel savings would exceed
700 million liters per year. These fuel savings produce a significant reduction in greenhouse
gases, saving nearly 650,000 tonnes per year. Thus, the project could show very real and lasting
results in the Philippines and in other Asian countries where 2-stroke emissions have a major role
in air quality.
The energy savings above are expected to produce economic benefits. For a typical
driver, the “fluid savings” from reducing the consumption of fuel and lubricating oil are
approximately $50 PhP/day ($1 USD). These savings should be compared to a typical driver’s
net income of approximately $250 Php5 ($5 USD) per day. A secondary economic benefit
involves the creation of small and medium enterprises. Heads, brackets, and wiring harnesses
will be manufactured in the Philippines. Kit installation centers will be established close to
concentrations of tricycles, and will employ mechanics and apprentices. Production estimates
have been established for the retrofit program. These estimates show that at the end of the first
year of Phase III, 5,000 - 10,000 retrofit conversions could be completed. Aggressive ramp-up is
proposed, with the potential for 500,000 conversions by the end of the third year.
Figure 3 – Students at Don Bosco Technical College watch the CSU team
install the prototype DI retrofit kit on an HD-III motorcycle

5.0 PROGRAM ORGANIZATION

For Phase I and Phase II of the project, Colorado State University is providing the
technical leadership. Swisscontact, a Philippines NGO, is undertaking project management
within the Philippines and Foundation for Sustainable Society (FSSI) provides a project oversight
role. Phase III of the project involves widespread conversion, requiring a “business-like”
operation which does not fall under the charters of these organizations. To prepare for Phase III,
Colorado State University has formed a standalone non-profit organization, EnviroFit International,
to support the widespread implementation program. It is anticipated that a corresponding
organization will be established in the Philippines during 2004 to organize and operate the retrofit
program. This will be a Philippines-based organization with local ownership and oversight. The
structure (for-profit vs. non-profit, etc) is currently being explored.
An element which is anticipated to be critical for the project’s success is the issue of
financing for the tricycle owners. While all efforts are being made to minimize the cost of the
system, the $10,000 – $12,500 Php ($200-$250 USD) cost of the kit may be an obstacle for the
tricycle owner. It should be noted that the fuel and oil savings seen from the system allow for a
one-year straight pay back of the cost of the kit, but few drivers would have the financial means to
invest this sum “up front”. If the initial capital can be provided through a micro-financing institution,
the potential for the success of the retrofit program will be greatly enhanced. It is anticipated that
the Phase III program may ultimately include micro-finance, partial subsidies, driver incentives,
and direct driver investment. All of these issues are currently being explored.
6.0 CONCLUSION

This paper describes an ongoing project to develop a direct injection retrofit kit for existing
two-stro ke motorcycles in Southeast Asia. Phase I of the project has been completed and a
functioning prototype is now in operation. Phase II will begin in early 2004 and will consist of
field-testing 10-15 units to verify reliability and emissions consistency. Phase III is scheduled for
2005 and will take the technology into production for widespread dissemination. Implementation
of this technology is expected to result in a 70%-90% reduction of HC, 50%-70% reduction in CO,
80% reduction in PM, 50% reduction in oil consumption, and 35% reduction in fuel consumption.
The estimated installed cost for the kit is $10,000 –$12500 Php ($200-$250USD). This cost is
approximately equal to the savings that an owner/operator would see in reduced fuel and oil
consumption over the course of a year. After that year of operation, the fluids cost savings of
approximately $50 Php per day ($1 USD) would equate to a 20% increase income contributing to
poverty alleviation. The infrastructure required to facilitate installation and service on the
retrofitted vehicles would provide new economic opportunities, providing jobs and training to
Filipino workers. Conservative estimates predict 5,000 kits in the first year, with 500,000 installed
by the end of the third year. If successful in the Philippines, this approach could be adapted to
other Southeast Asian countries that are impacted by high populations of vehicles with two-stroke
cycle engines.

1
Marco Nuti. "Emissions from Two-Stroke Engines," Society of Automotive Engineers, 1998, pg 85.
2
Sam Leighton, and Stev e Ahern, "Fuel economy advantages on Indian 2-stroke and 4-stroke motorcycles
f itted with direct f uel injection," SAE Technical Paper No. 2003-26-0019.
3
Mark Archer, and Greg Bell, "Advanced Electronic Fuel Injection Systems – An Emissions Solution for both
2- and 4-stroke Small Vehicle Engines," SAE Technical Paper No. 2001-01-0010.
4
Colorado State Univ ersity - Bry an Willson, Jessica Rupp, Michael Duncan, Dan Mastbergen, Tim Bauer,
Nathan Lorenz, Steve Nelson, Nick Peterson, Bryan Wedryk, Trav is Mathis, "Colorado State Univ ersity –
Clean Snowmobile Challenge 2002," SAE Technical Paper No. 2002-01-2758.
5
Standards Development for the Local MC/TC Sector-Third Quarter Progress Report to DOST-PCIERD
U.P. National Center f or Transportation Studies Foundation Inc. (October-December 2002)